Odd Channel Difference

Post any questions you may have about our current range of oscilloscopes
Post Reply
Posts: 0
Joined: Thu Jan 09, 2014 3:26 pm

Odd Channel Difference

Post by Graybags » Fri Feb 05, 2016 2:33 pm

I've noticed an odd difference between channels - it is NOT an issue for me, unless it is an indication of something else I've not yet discovered. Nevertheless, I'm hoping someone may be able to offer an explanation.

I'm using a 2205 MSO in analog mode only. Picoscope version

While 'zeroing' the probes I've noticed different behaviour between ChA and ChB despite (TTBOMK) them being set the same, i.e. Probe x10, Resolution 8 bits Input Range of +/- 5V. Both probe tips are connected to their own ground clip.

Chan A appears to be much more 'noisy' than Chan B - not in the peak2peak value of the noise (both channels ~50mV which I believe is normal and due to digitization) but in the 'number of transitions' (I'm not sure what to call this, frequency seems wrong). For example at 1uS/div Chan A has ~200 mini spikes while Chan B has 5. I've tried physically swapping the probes, Chan A remains more noisy. I've also readjusted the frequency compensation of both probes.
I've attached a pdf of a couple of traces (with both axis scaled to x10).
Two Chan Zeros.pdf
(90.91 KiB) Downloaded 182 times

Site Admin
Site Admin
Posts: 186
Joined: Thu May 25, 2006 8:02 am

Re: Odd Channel Difference

Post by alan » Sat Feb 06, 2016 2:08 pm


Agree this looks odd but it is normal - let me explain.

You are on a range which spans from -5V to +5V (a 10V span). The scope is an 8 bit scope which means the ADC can record 256 different levels. 10,000 mV / 256 = 39mV per step.

You have zoomed in vertically by a factor of 10 and both traces are jumping up by 39mV which is one least significant bit of noise - the smallest step possible. Many other scopes will show several counts of noise so this is actually really good for an 8 bit scope.

0V often falls between 2 different ADC values and will vary a bit from channel to channel - this is taken care of when the unit goes through final test and calibration. If one channel ends up with 0V between 2 ADC counts the small amount of random noise will mean the channel jumps between the values - you are seeing this on channel A.

If however you see a trace which is mostly staying at one ADC count, like channel B this is because 0V ends up nearer one ADC level than the next.

In the end both channels are exhibiting 1 LSB of noise - its just that one stays mostly on one ADC value with the odd jump up and the other jumps between 2 values. All normal and as mentioned above only having 1 LSB of noise is actually rather impressive.

Post Reply